Mood Stabilizers in Prophylaxis: Should we ever Discontinue? E. Timuçin Oral, MD Bakırköy Prof Mazhar Osman Research and Training Hospital for Psychiatric & Neurological Diseases İstanbul / Turkey
Treatment of Bipolar Disorder
Cross sectional Longitudinal
Cross sectional
Targets of Prophylactic Treatment in Bipolar Disorder • • • •
Reducing mortality Reducing the duration of “sick period” Preventing relapse & recurrence Improving functioning & QoL
The Consequences of Suddenly Stopping Psychotropic Medication in Bipolar Disorder • Li discontinuation increases the risk of mania • Abrupt discontinuation (<2 wk) carries greater risk than gradual (2–4 wk), no optimal period • Discontinuation with all major ADs: hypomania has also been reported • Further research in BP is required for the discontinuation effects of AC & AP drugs • Discontinuation clinically important in choosing treatment for non-compliant, pregnancy and in stopping treatment after a long remission
Defining Discontinuation Discontinuation syndrome ? Rebound phenomenon ? Withdrawal effect ? – withdrawal implies dependence (tolerance ? Addiction ?) – rebound requires a temporary rise in symptom frequency to a higher level than previously experienced
“The concept remained unproven” Schou,1993
Treatment Guidelines for BPD of TPD Algorithm for “Maintenance Treatment” E.T.Oral, 2003 (Vahip S, Yazici O: Eds) 2nd or later episodes
1st episode
Not taking MS
Severity of the episode
Taking MS
Quality of life Positive family history The choice of the patient
Cont. MS started at acute episode or
Start MS (Li)
Change MS
Decision: YES Decision: NO
Recurrence or partial/minimal response
Full remission
Taper down MS and STOP
Decision
Combine 2 MSs
Continue treatment Alternatives
Lithium Maintenance Therapy Recommended Ranges Schou & Baastrup,1967 APA Task Force, 1975, Prien & Caffey, 1977 Schou, 1984
0.6 to 1.5 mEq/L 0.7 to 1.2 mEq/L 0.6 to 0.8 mEq/L 0.6 to 0.8 mEq/L < 0.4 mEq/L, risk of relapse
NIH Consensus Panel, 1985 Schou, 1997 0.6 to 0.8 mEq/L Recommends 0.5 to 0.8 mEq/L, but < 0.5 mEq/L in elderly 0.8 mEq/L in younger “Changes in serum lithium levels as small as 0.1 to 0.2 mmol/L, upward and downward may improve patients’ quality of life
Retrospective Studies that Found a Relationship Between Lithium Level and Response Study
N
Baastrup & Schou 1967
88
Prien & Caffey 1976
32
Comments 80% responded at1.0 0.5mEq/L
> 0.1 mEq/L were no more effective1.0 than mEq/L 0.8 0.7 mEq/L were no more effective than Venkoba Rao & 28 12 had good responses at >0.8 mEq/L. 6 were at placebo. Hariharasubramanian1978 low levels,” “very James et al. 1980 100 Respondersmean level: 0.69 mEq/L Nonresponders mean level: 0.58mEq/L Sarantidis & Waters 1981 46 Fair response (n=6) 0.63 Excellent response (n=31) 0.74 mEq/L No respo nse (n=9) 0.79 mEq/L Jann et al. 1982 30 12 switched (once = 0.66 mEq/L, twice=0.48 mE did not switch (0.77 mEq/L). Peselow et al. 1982 116 Significantly higher lithium level during euthymia. Sashidharan et al. 198253 (41 BP, good outcom e < 0.9 mEq/L than >0.9 mEq/L. 12 UP) Yang 1985 101 2% responded at0.49 0.4 mEq/L, 78% at0.79 0.5 mEq/L, and 20% at1.09 0.8 mEq/L.
Retrospective Studies that Did Not Find a Lithium Level-Response Relationship Study Coppen et al. 1971 Persson 1973 Dunner & Fieve 1974
N
Levels (mEq/L)
65
0.73-1.23
No relationship
53 (23BP, 30 UP)
average, 0.7
55
0.7 - 1.2
No significant difference in relapse <0.6 and >0.6 rate between No differen cebetween responders & nonresponders
Vinarova & Vinar 1984
39 responders 17 nonresponders Goodnick & Fieve 1985 44 Lokkegaard et al. 1985 Maj et al. 1985 Smigan 1985
Comment
153 43 responders 25 nonresponders 49 responders 23 nonresponders
0.76 0.73 0.58vs0.87
No signifcant difference in mood symptoms or side effects 0.73, 0.77, 0.80 No significant difference among 3 groups mean: 0.65 mean: 0.66 mean: 0.63 mean: 0.61
Standard vs. Low Lithium Levels • Lower range associated with a 2.6-times increased risk of relapse • Cycling episodes - 11 patients on low level vs 2 on standard level • Three or more episodes means ? effect of lithium at either level • Median lithium levels: 0.54 + 0.12 mEq/L (low range) and 0.83 + 0.11 mEq/L (standard range)
Patients receiving lithium doses that achieved standard serum levels (0.8 to 1.0 mEq/L) had better psychosocial functioning than those receiving doses that achieved low serum levels (0.4 to 0.6 mEq/L); this effect was partially but not wholly mediated through relapse prevention. Solomon et al. Am J Psychiatry 1996;153:1301-1307
• Relapse rate 50% within 5 months 90% within 1.5 year Suppes 1991
• 10-15% of patients, well for sustained periods of time, who chose to discontinue Li and experienced relapse, failed to re-respond Post 1992; Bauer 1994; Koukopulos 1995; Maj 1995; Tondo 1997
• Abrupt decreases in Li (even in therapeutic range) were a powerful predictor of relapse Perlis 2002
QUESTION? • After what duration of Li treatment do its benefits outweigh the possible risk of recurrence on withdrawal? • In discontinuing Li after 6 months, the risk was brought forward by 20 months, after 18 months it was brought forward by ~7 months. ANSWER: • “Li should not be used for <2 years” GM Goodwin, Br J Psychiatry 1994
The risk of relapse post-Li discontinuation after (A) 6 months and (B) 18 months. The control risk in first-episode bipolar patients not treated with Li is from Mander (© The Royal College of Psychiatrists, 1994)
• BP women discontinued Li maintenance • 42 pregnant X 59 nongravid followed during pregnancy + postpartum 24 wk & equivalent times • Discontinuation: Rapid (<14) X Gradual (15 - 30) • First 40 wk → little difference in recurrence (overall 55.4 %) • Postpartum → Pregnant : Nonpregnant = 2.9
Viguera AC, et al. Risk of recurrence of bipolar disorder in pregnant and nonpregnant women after discontinuing lithium maintenance. Am J Psychiatry February 2000;157: 179-84
• 3/9 women receiving Li during pregnancy had recurrence in two weeks after delivery • In 64-wk after Li discontinuation, recurrences were common (85.7 % in pregnant/postpartum vs 67.8 % in nonpregnant) • 30% of 20 pregnant X 76% of 25 nonpregnant women remained stable 6 months more • During first 40 wk recurrence risk was 2.5 times shorter in rapid discontinuation. Viguera AC, et al. Risk of recurrence of bipolar disorder in pregnant and nonpregnant women after discontinuing lithium maintenance. Am J Psychiatry February 2000;157: 179-84
P<0.0001
Patient (%)
AD
Viguera et al. 2001
• Postpartum period brings great risk for women • Postpartum recurrence can be reduced by Li maintenance during pregnancy • Gradually discontinuing lithium limits the recurrence risk during the first 40 wk
Viguera AC, et al. Risk of recurrence of bipolar disorder in pregnant and nonpregnant women after discontinuing lithium maintenance. Am J Psychiatry February 2000;157: 179-84
• 78 BPs Rapid Li discontinuation (<14 days) Gradual discontinuation (15-30 days) • Time to recurrence x 5.6 gradual (14 m) > rapid discontinuation (2.5 m) Baldessarini RJ et al, reduced morbidity after gradual discontinuation of lityum treatment for bipolar I and II disorder: a replication study.Am J Psychiatry1997;154)
• Rapid discontinuation (<7) = 47% recurrence risk • Special attention in rapid discontinuation needed Baldessarini RJ, Tondo L: Recurrence risk in bipolar manic-depressive disorders after discontinuing lithium maintenance treatment: an overview. Clin Drug Investigation 1998; 15:337-351
Controlled Lithium Discontinuation in Bipolar Patients with Good Response to Long-term Lithium Prophylaxis. O Yazıcı, K Kora, A Polat, M Şaylan Istanbul Faculty of Medicine, University of Istanbul J Affect Disord. 2004 Jun;80(2-3):269-71
Discontinuation of Long-term Treatment • BP-I (DSM-IV) • Discontinuation of a >5 year successful, long-term lithium prophylaxis • Definite good response • Controlled cessation • Prospective follow-up • 32 patients (16 M, 16 F)
Discontinuation of Short-term Treatment • BP-I (DSM-IV) • Discontinuation of a <5 (1,5 -4,5) year successful, long-term lithium prophylaxis • Definite good response • Cessation • Retrospective evaluation • 30 patients (11 M, 19 F)
Clinical Characteristics During Discontinuation Duration of illness (yr) Total episodes Li usage (yr) Hospitalizations Age of Li use First episode ratio (%) Mania Depression Psychotic episodes (%)
LT 15 10 6.2 2 35
ST 9 5 2.7 1.4 33
59 41 25
80 20 33
p <0.002 <0.005 <0.0001 NS NS NS NS
Clinical Characteristics During Discontinuation Age of Discontinuation Psychiatrists approval (%) Feeling well enough (%) Rate of Discontinuation (%) Rapid Slow
LT 41 100 88 72 28
ST 36
p NS
37 <0.0001 78 NS NS 73 27
Recurrence Rates (%) One Week 6 Months One Year
LT 7 32 62
ST 10 26 66
NS
30% of the patients had recurrence in one week, 50% in 6 months and more than 60% at the end of first year.
Results • • •
No differences between two groups by means of sociodemographic variables There were no BP-II patients in the shortterm treatment group Total duration of illness, number of episodes, duration of lithium usage and approval of discontinuation by physician is hisgher in LT group
Results • •
•
Episode type, severity latency and rate of recurrence are equal in both groups Being in prophylactic treatment more than 5 years did not make any difference in recurrence rate when Li was discontinued
Discontinuation of Valproate in BP Maintenance: One-year follow-up Toksoy OM, Erten E, Verimli A, Oral ET Bakırköy State Teaching and Research Hospital for Psychiatry & Neurology / Istanbul
European Neuropsychopharmacology 12-S212 Suppl. 3 2002
Objective The risk of a new episode and predictors of treatment were investigated in BP patients being in remission at least for 6 months after the discontinuation of VPA which was started as an adjuvant agent to Lithium during the acute episode and continued during the maintenance period of treatment.
INCLUSION CRITERIA
EXCLUSION CRITERIA
DSM-IV BP I 6 m remission
Abnormal EEG Epilepsia
Li+VPA in prevention Age 18-65 y
Hx Org. Brain Disease Not responded sole Li
66 remitted BP (34 female 32 male) in age 18-65 VPA discontinued abruptly (3 days) or slowly (15 days) 33 patients in each group Followed for 1 year
Li
Li + VPA + AP
Li + VPA
Acute Period
Main.Period
2-12 weeks
6 months
(VPA tapered in 3 d.)
Follow-up period After VPA Discontinuation Li (VPA tapered in 15 d.)
12 months
Mean age: 32.44 ±9.92 (18-58) Male 32 (48.5%) Female 34 (51.5%) BP history in first degree relatives 40.9 % No correlation between life events, occupation, marital status and episodes Duration of Treatment Li ONLY 27 (42.9%) 17 (27 %) 19 (30.2%)
Li+VPA < 2 yrs 2 –5 yrs 5 > yrs
33 (54.5 %) 17 (25.8 %) 13 (19.7 %)
<1 yr 1-2 yr >2 yr
Ilness Characteristics • Age of onset • Total # of episodes • Type of first episode •Manic •Depression • Mixed
24,05 ± 8,4 4,24 ± 4,2 75,8 % 22,7 % 1,5 %
(14-49) (1-20)
Ilness Characteristics II • Psychotic features
75%
•Mood Congruent 53% •Mood Incongruent 22% • Rapid-cycling
• Seasonality • Premenstruel onset • Postpartum onset
3% 37,9% 5,8% 5,8%
Results I Episode in the first year :28 patients (43.1%) (11 female, 17 male) Duration of a new episode:149 ± 91.5 (8-300 dy) Type of the new episode: Hypomania-mania 27.7% Depression 13.8% Mixed
Results II Type of Discontinuation / New Episode Abruptly Gradual
new episode new episode
40.6 % 45.5 %
* NS
Patients hospitalized: 17.8 % *** (All in abrupt group) # of manic patients: # of depressive patients: 1
4
Blood Levels at Discontinuation • Lithium • 6% •79.8 % •14.2 % • Valproate •54 % •46 %
0.6 mEq / lt > 0.6 – 0.8 mEq / It 0.8 mEq / It <
45 – 60 mg / lt 60 – 80 mg / lt
Results Risk of a new episode is similar between slow and abrupt discontinuation groups. In the second group, episodes seemed to be more severe. Predictors of Relapse – Prior history of suicide – Alcohol abuse
(p<0.01) (p<0.01)
– Male gender – Subsyndromal symptoms
(p<0.05) (p<0.01)
Results • After VPA discontinuation 56.9 % of the patients were still in remission. • Episodes were seen in the first 6 month of discontinuation in 22 / 28. • In abrupt discontinuation group, episodes were more severe and lasted longer. (Duration of the illness was higher in this group which may also be a reason for severity )
Li Rebound Phenomena ? • 2 retrospective, 4 RCT-DB, 2 RCT-CO, 1 meta-analysis • 29 rapid Li stoppers had 51.7% recurrence in 3 months Franks et al. The consequnces of suddenly stopping Psychotropic Medication in Bipolar Disorder, Clinical Approaches in Bipolar Disorders 2005; 4:3)
• Meta-analysis on 14 Li discontinuation studies (n=257) • Time to 50% relapse was 2.5 months. Mania (med: 2.5 month) > Depression (med: 6 month) • New episodes / month = X 27.9 between patients continuing and discontinuing Li. Suppes et al, Arch Gen Psychiatry 1991
• Li discontinuation may increase the risk of suicide • 185/310 BP on long-term Li, discontinued treatment • Risk of suicidal acts was X 7.5 greater Tondo,L. Acta Psychiatr Scand, 2001
Anticonvulsant Discontinuation • AC & AP are now widely used to treat BP • Few studies investigated the discontinuation • One retrospective study of 6 BP discontinued CBZ did not demonstrate a rebound effect. Following 3 months none of them experienced a manic episode, one developed depression Macritchie KA, J Psychopharmacol 2000
Antipsychotic Discontinuation BP patients in remission with OLZ Continuing patients (n=225) X Rapid stoppers (n=136) Recurrence Rate Time
OLZ 46.7% 174 days
PBO 80.1% 22 days Tohen, Arch Gen Psychiatry 2005